Veronique Francois, Project Manager, Pistoia Alliance
Increasingly, regulators are looking to shift regulatory submissions from PDF to data streams. Preclinical data is no exception. Reviewers want standardized, machine-readable data, and we’re working to make that happen.
Who should read this
If you build, run, or analyze in vitro assays using traditional or New Approach Methodology (NAM) models to test the activity and safety of a drug candidate, or if you move the data they generate, you’ll be interested in this article.
The problem
The same target undergoes preclinical testing in many different ways, generating data in various formats. Submitted to regulatory authorities, reviewers slog through that diversity and can’t compare across studies. Review delays and frustration will only grow as opportunities to reduce animal use let NAMs gain traction and be admitted for submissions.
What regulators are moving toward, and how we’re helping
Regulators are increasingly aiming to streamline the submission of preclinical data by automating processing and standardizing data. That’s why our projects focused on standardizing in vitro analyses — In Vitro NAM Data Standards and In Vitro Pharmacology — are in active exchange with regulators. As we work toward standardized data structures, harmonized terminology, and minimum metadata requirements, we are incorporating guidance from regulators like the FDA on submission templates. The aim is a common way to submit in vitro data that people and machines can read alike.
In vitro NAM: concrete, and tested on real safety data
We are working on a standardized data structure that builds on SDTM to capture assay metadata, study design, and study results. The structure has been tested and validated on real data from human iPSC-derived cardiomyocyte assays and hepatocyte assays. These assays cover two of the most common safety failure modes: cardiotoxicity and hepatotoxicity.
But there is more work to be done. As Stefano Gaburro pointed out in his article “Regulate NAMs Data Like SEND, Then Go Beyond It,” we need to go beyond existing structures because NAMs carry context (cell source, donor genetics, device geometry, media) not found in the assay types SDTM was designed for.
As we build harmonized annotation and terminology and move toward generating a dataset suitable for downstream AI/ML integration, we work alongside the FDA. We are also looking to expand that collaboration to include standards and validation organizations, ensuring the long-term maintenance of the standards we create.
In Vitro Pharmacology: the first step taken, more to come
Our In Vitro Pharmacology project aims to develop standardized data formats and tools for in vitro bioassays. Our exchange with regulators, including the FDA, has narrowed our focus to a concrete goal: a standardized template for data submission. Getting there, however, requires establishing a unified data framework based on the assay types in use and existing ontologies for regulatory submissions.
Because the assays are molecular (binding affinity, protein-protein interaction, enzyme activity, and the like), they don’t fit existing data models. So, we are building these standards based on our experience from the DataFAIRy Bioassay Annotation project, on the BioAssay Ontology (BAO), and by capturing standardized assay descriptions. To that end, we created a repository for in vitro pharmacology assay descriptions, which will eventually link to data for submission in the template agreed upon with regulatory authorities.
In the short term, the repository will produce computer-readable data that improves traceability, connectivity, and analytical insight. You can already explore the repository here.
And this is all part of a bigger picture
These two in vitro data standards projects are part of three standards development projects for preclinical data at the Pistoia Alliance. We run them in parallel, steering each toward a standards architecture that will make preclinical data interoperable. For example, a related effort uses proven toxicology standards (SDTM) as a springboard for in vivo efficacy data standards.
For now, our focus is data exchange within the three domains. The vision, ultimately, is a more integrated, interoperable data ecosystem for preclinical data.
If you work on preclinical investigation of drug candidates, this is where the standards for your data are being written. Join us to make them useful for everyone.
Get involved
The data structures, terminology, and templates are being defined now, in active exchange with regulators. This is the point at which new contributors shape the outcome most. If your organization builds, runs, or submits in vitro data, this is the moment to help write the standards.
Join or learn more: In Vitro NAM Data Standards and In Vitro Pharmacology
About the Pistoia Alliance preclinical data-standards projects
The Pistoia Alliance is a global, not-for-profit alliance that lowers barriers to innovation in life science R&D through pre-competitive collaboration. In Vitro NAM Data Standards and In Vitro Pharmacology are two of three preclinical projects, alongside In Vivo Efficacy Data Standards, building the standards that let preclinical data be exchanged, compared, and interoperable.
Consistent, well-structured data builds on a shared infrastructure the whole industry benefits from building once, together. That’s why this work happens in a pre-competitive setting: no single company has to solve it alone.
Your turn
If you’ve submitted in vitro assay data to a regulator, where did the format cost you the most? Was it reconciling terminology, mapping to a template, or something the reviewer couldn’t compare?